TELSTRA CORPORATION LIMITED submission_9.pdf · Telstra Corporation’s Response to the Commission’s Domestic Mobile Terminating Access Service Declaration Inquiry – Report of

Telstra Corporation’s Response to the Commission’s Domestic Mobile Terminating Access Service Declaration Inquiry – Report of the ACCC’s Draft Decision

TELSTRA CORPORATION LIMITED (ABN 33 051 775 556) |

PAGE 1

TELSTRA CORPORATION LIMITED

Response to the ACCC’s Mobile Terminating Access Service final access determination –Draft Decision

Public version

15 June 2015



PAGE 2

CONTENTS

01 EXECUTIVE SUMMARY 3

02 INTRODUCTION 6

03 MOBILE VOICE TERMINATION SERVICES 7

3.1. TSLRIC+ pricing framework 7

3.2. Mobile-to-mobile and fixed-to-mobile termination 8

3.3. International benchmarking 8

3.3.1. Selection criteria for benchmark countries 9

3.3.2. Adjustment factors 11

3.4. WIK-Consult international benchmarking approach 24

3.5. Draft mobile voice termination rates for 2016-2019 24

04 SMS TERMINATION SERVICES 26

4.1. A regulated price for SMS termination 26

4.2. Approach to pricing SMS termination 27

4.3. Draft SMS termination rates for 2016-2019 30

05 FIXED-TO-MOBILE PASS THROUGH 31

5.1. A mandated FTM pass-through mechanism will distort competition 31

5.2. Reductions in MTAS rates have been passed through to end-users 31

06 DURATION OF REGULATED TERMS AND CONDITIONS 33



PAGE 3

01 EXECUTIVE SUMMARY

1. Telstra welcomes the opportunity to provide this submission in response to the Australian Competition and Consumer Commission’s (ACCC) Mobile Terminating Access Service (MTAS) Final Access Determination (FAD) – Draft Decision.

2. The Draft Decision sets out the ACCC’s preliminary view on primary price terms for the new MTAS

FAD. The primary price terms include the regulated prices for the mobile voice termination service and the short messaging service (SMS) termination service.

3. Consumers in Australia continue to benefit from a thriving and dynamic mobiles market with high

levels of investment and improvements in the customer experience. Pricing of mobile services continues to change in response to competition and consumer demand. All mobile network operators (MNOs) have recently announced changes to mobile plans incorporating – for example - higher data allowances, ‘bonus’ inclusions and the ability to further tailor mobile plans to suit individual customer requirements. The ACCC has separately noted an increase in price competition, as well as numerous other developments such as a competitive focus on data services, investment in 4G networks and customer strategies like Telstra’s real time data alerts which are aimed at helping customers avoid bill shock.

1 In this context, the approach the ACCC is

currently taking to not regulating the mobiles sector except in relation to mobile interconnection between carriers is clearly resulting in outcomes which are in the long term interests of end users (LTIE).

4. Telstra supports the ACCC’s approach to determining the mobile voice termination rate using a

TSLRIC+ pricing methodology. In contrast to other pricing methodologies, a TSLRIC+ framework will promote the LTIE as it is a cost-based methodology that allows a return on efficiently invested capital and the recovery of efficient common costs. Telstra also agrees with the draft decision to retain symmetric rates for fixed-to-mobile (FTM) and mobile-to-mobile (MTM) termination services. Telstra considers that the application of the same price to FTM and MTM termination services is cost-reflective and avoids the risk of inefficient arbitrage.

5. Telstra also agrees with the ACCC draft decision not to adjust the mobile voice termination rate to

take account of Voice-over-LTE (VoLTE). Telstra considers that the timing of commercial VoLTE services is not yet certain due to the need to develop industry standards and ensure the availability of compatible handsets. In any event, while MNOs may launch VoLTE services during the next FAD period, Telstra considers it unlikely that deployment will occur to the extent that it impacts on the efficient costs of providing mobile voice termination. At this point in time, any adjustments to mobile voice termination rates to take account of VoLTE would be speculative and risk regulatory error.

6. In the absence of the development of a new cost model for MTAS, Telstra agrees with the ACCC that a pragmatic and balanced approach to pricing mobile voice termination services is the use of appropriate international benchmarks. Telstra considers that the objective of international benchmarking should be to – so far as possible – replicate the outcomes of a cost model approach. This requires that the countries included in the benchmark sample be comparable to Australia and that benchmark costs be further adjusted to appropriately reflect the Australian context including for the different cost drivers that arise due to, for example, differences in population density and distance.

1 ACCC, Telecommunications Report 2013-14: https://www.accc.gov.au/publications/accc-telecommunications-

report/accc-telecommunications-report-2013-14



PAGE 4

7. Telstra engaged Network Strategies to review the international benchmarking undertaken by WIK-Consult on behalf of the ACCC. This review identified a number of methodological errors in the calculations made by WIK-Consult as well as broader issues such as the failure to use the latest version of benchmark cost models that are available. However, the lack of transparency of WIK-Consult’s approach and insufficient supporting analysis make it very difficult to provide appropriate or accurate feedback on the international benchmarking approach and the mobile voice termination rate proposed by the ACCC.

8. Nevertheless Telstra considers that there are significant identifiable shortcomings with the

approach taken by WIK-Consult including, but not limited to:

a. Failure to adjust the benchmark sample to ensure that the countries included are comparable to Australia despite clear differences in relation to population density, land area and size of the mobile network;

b. Absence of adjustments to reflect the considerable differences in population density, which is likely to significantly understate the cost of deploying and maintaining mobile networks in Australia

These factors, in addition to the numerous other issues identified by Network Strategies and the lack of transparency regarding the approach taken by WIK-Consult, mean that the regulated rate proposed by the ACCC is in Telstra’s view unreliable. This is supported by the fact that, as shown in Figure 1 below, the proposed rate of 1.61 cents per minute (cpm) is the lowest TSLRIC+ voice termination rate amongst a sample of 30 international jurisdictions.

Figure 1 International mobile voice termination rates

0 2 4 6 8 10 12

Romania

Malta

Denmark

Norway

France

Sweden

Austria

United Kingdom

WIK benchmark

Netherlands

Belgium

Italy

Hungary

Spain

Cyprus

Croatia

Slovenia

Czech Republic

Mexico

Portugal

Germany

Turkey

Poland

Slovakia

Kenya

Bulgaria

Malaysia

South Africa

Nigeria

Tanzania

Indonesia

Voice termination rate (AUD cents)

Pure LRIC

LRIC+



PAGE 5

9. The implication that the TSLRIC+ based estimate of the efficient cost of providing mobile voice

termination services in Australia is only marginally above the Pure LRIC based estimate for the United Kingdom (UK) suggests that the approach taken by WIK-Consult is flawed. Further the errors identified in the approach and calculations undertaken by WIK-Consult fail to provide the evidence required by the ACCC to support the introduction of a new primary price term for mobile voice termination. Telstra considers that the ACCC must reassess its acceptance of the mobile voice termination rate calculated by WIK-Consult and that this should be done in consultation with industry to ensure accuracy and appropriateness of a primary price that reflects the efficient cost of providing mobile voice termination in Australia.

10. Telstra continues to consider that there is no requirement to set a regulated price for SMS termination. The competitive nature of the SMS market in Australia, including the use of alternative messaging services and platforms, makes the application of regulatory assumptions used to set primary price terms for SMS termination highly uncertain. The substitutability of alternative messaging services and platforms is supported by the ongoing growth in smartphone penetration across all market segments. Further, the limited number of jurisdictions (comparable or otherwise) from which to derive international benchmarks for SMS termination, as noted by the ACCC, suggests that regulated pricing of SMS termination is unnecessary.

11. Notwithstanding Telstra’s view that there is no need to set a regulated price for SMS termination,

Telstra considers that, while the approach taken by the ACCC in the Draft Decision is sound in principle, the practical application of the approach by WIK-Consult is flawed. In particular, the use of international benchmarks – with unsupported assumptions and adjustments - to estimate the cost of investments in SMS Centres in Australia where local information is available is unnecessary and increases the risk of regulatory error. Telstra’s position is that the ACCC should work with the MNOs to provide consistent information that reflects the efficient cost of providing SMS termination in Australia. Further Telstra considers that, given the distinct nature of Application to Person (A2P) SMS, applying price regulation to this growing market risks inhibiting further development and innovation, which would not be in the LTIE

12. Telstra supports the ACCC’s draft decision not to include a mandated FTM pass-through

mechanism in the FAD. This decision has been largely made on the basis of ACCC analysis which provides evidence that Telstra has passed on a significant portion of savings from reductions in the mobile voice termination rate. This confirms analysis previously undertaken by Telstra. Telstra considers that the ACCC’s draft decision is in the LTIE as it ensures that service providers will continue to have the flexibility to determine how cost savings should be passed on to customers across the bundle of fixed services. This further ensures that customers benefit from cost reductions where they are most highly valued. Accordingly, Telstra agrees with the ACCC that a mandatory pass-through mechanism may “...damage economic efficiency and undermine the promotion of competition and therefore may not promote the LTIE.”

13. Telstra agrees with the ACCC draft decision that the price and non-price terms of the MTAS FAD

should expire at the same time as the current MTAS declaration on 30 June 2019. The ACCC notes that the commercial launch of VoLTE may warrant a review of the primary price terms of the FAD before its expiry should there be evidence that the regulated termination rates no longer reflect the efficient costs of service provision. As noted above, Telstra considers that the timing of VoLTE deployment is not yet certain and is unlikely to impact on the efficient cost of providing the MTAS during the FAD period. Further Telstra considers that the ACCC should place priority on regulatory certainty at a time when the industry as a whole is going through a period of transition.



PAGE 6

02 INTRODUCTION 14. This submission is structured as follows:

a. Section 3 sets out Telstra’s views on the Commission’s draft decision on price terms for mobile voice termination

b. Section 4 sets out Telstra’s view on the Commission’s draft decision on price terms for SMS termination services

c. Section 5 sets out Telstra’s view on the Commission’s draft decision relating to fixed-to-mobile pass through

d. Section 6 provides Telstra’s position on the duration of regulated terms and conditions

15. Telstra may make further submissions to the ACCC once it has had an opportunity to review industry responses to the Draft Decision.



PAGE 7

03 MOBILE VOICE TERMINATION SERVICES

16. The Draft Decision sets out the ACCC’s position on the pricing approach to mobile voice termination services. Specifically that:

a. TSLRIC+ remains the most appropriate pricing framework.

b. Symmetric rates should apply to mobile-to-mobile and fixed-to-mobile termination.

c. International benchmarking is the most appropriate pricing methodology.

d. The benchmarks are based on a TSLRIC+ framework with adjustments to take account of Australia-specific factors.

e. A rate of 1.61 cents per minute (cpm) will apply from 1 January 2016.

Telstra’s position on the above is set out below.

3.1. TSLRIC+ pricing framework

17. The Draft Decision sets out the ACCC’s conclusion that TSLRIC+ remains the most appropriate

pricing framework on the basis that it ensures that MNOs are appropriately compensated for the provision of the MTAS, promotes allocative efficiency, protects the legitimate business interests of the MNOs and provides sufficient incentives for MNOs to maintain and invest in the infrastructure necessary to provide the MTAS.

18. Telstra agrees with the ACCC’s conclusion – a TSLRIC+ based price for the MTAS will best promote the LTIE as it provides an access provider with an expectation of a return on efficiently invested capital and the recovery of efficiency incurred common costs. Telstra has previously submitted that TSLRIC+ will promote the LTIE by:

a. Encouraging the efficient use of infrastructure and investment in mobile infrastructure;

b. Continuing to promote competition;

c. Promoting the legitimate interests of an efficient access provider; and

d. Providing for regulatory certainty.

19. The mobile sector is characterised by high levels of innovation and corresponding levels of investment. All MNOs are continuing to invest in improving the coverage and quality of 4G networks as well as ensuring that mobile services continue to meet the evolving needs of mobile customers, particularly with respect to the provision of data services. The ACCC has separately acknowledged that price competition in the sector has recently picked up, alongside other developments such as Telstra’s real time data alerts, which help customers to avoid bill shock.

2

The high level of competition is also demonstrated by introduction of new mobile plans by all MNOs which provide, among other things, flexibility in data sharing and ‘bonus’ inclusions.

3 A TSLRIC+

2 ACCC, Telecommunications Report 2013-14, pages 31-32.

3 Optus released new prepaid mobile plans in January 2015 which include extra data, unlimited weekend calls and

unlimited texts. In March 2015, Optus also offered new and re-contracting customers on certain mobile plans a six month subscription to Netflix. In May 2015, Telstra released Go Mobile plans which allow data to be shared between devices and include a 6-month Presto subscription, 12-month AFL Live Pass or 12-month NRL Digital



PAGE 8

methodology creates incentives for ongoing investment by affording access providers an opportunity to recover relevant costs, while at the same time promoting competition by ensuring efficient pricing outcomes.

20. TSLRIC+ also promotes the legitimate business interests of an efficient access provider through

the inclusion of a return on efficiently invested capital and a return of efficiently incurred costs. Further, in the presence of the strong market outcomes currently observed in the mobiles sector, there is no rationale for the ACCC to depart from the well-established TSLRIC+ methodology to determine the efficient price of the MTAS in Australia. The continued use of TSLRIC+ provides for regulatory certainty and reduces any risk of regulatory error from the application of an alternate methodology.

21. The ACCC has identified a number of shortcomings associated with other potential methodologies

for pricing the MTAS; including a pure LRIC framework, building block model and setting a mobile voice termination rate relative to a fixed termination rate. Telstra agrees with the conclusions in the Draft Decision that these approaches are not appropriate for pricing the mobile voice termination service. In particular, Telstra considers that a Pure LRIC approach is not in the LTIE as it would price the MTAS below full cost recovery. This will result in an inefficient use of infrastructure and investment, distort competitive outcomes and does not support the legitimate business interests of access providers. Further detail on the inappropriateness of a Pure LRIC approach has been set out in previous submissions.

4

3.2. Mobile-to-mobile and fixed-to-mobile termination

22. Telstra supports the ACCC draft decision that both MTM and FTM voice termination rates should

be set at the same price using the same methodology. Telstra considers that this approach is appropriate. As set out in previous submissions

5, asymmetric pricing – whether between services

or between MNOs – raises the risk of arbitrage and, in the absence of calling externalities, will result in inefficiencies that are not in the LTIE. Further, as noted by the ACCC, the MTM and FTM termination services are technically identical, using the same infrastructure and costing the same amount to provide. Retaining symmetric rates therefore ensures that the mobile voice termination rate is consistent with the ACCC’s cost-based approach to the MTAS.

3.3. International benchmarking

23. Telstra considers that a forward-looking bottom-up (FLBU) model is the best practice approach to setting a TSLRIC+ price for mobile voice termination services. This approach is consistent with the previous WIK model used by the ACCC, as well as that used by many other jurisdictions worldwide. Given that the WIK model is now out-dated, ideally the ACCC would develop a new TSLRIC+ model. However, as noted by the ACCC in the Draft Decision, the development of a new cost model would be both time- and resource-intensive as well as cause delays to the FAD process. Accordingly, Telstra agrees with the ACCC that a more pragmatic approach to pricing mobile voice termination services in Australia is the use of international benchmarking. However, Telstra considers that international benchmarking is only an appropriate approach insofar as it replicates (as much as possible) the efficient outcomes of a TSLRIC+ model through the selection of an relevant benchmark set and by making adjustments to account for the Australian context.

24. The ACCC engaged WIK-Consult to undertake an international benchmarking study to estimate the cost of mobile voice termination in Australia. The results of the benchmarking study were

Pass. Also in May 2015, Vodafone re-introduced its ‘double data’ promotion across a range of mobile plans with some plans including a choice of entertainment subscription to Stan, Spotify Premium or The Age/SMH. 4 Telstra Corporation’s Response to the Commission’s Mobile Terminating Access – Final Access Determination

Discussion Paper, 5 September 2014. 5 Ibid.



PAGE 9

summarised in the Draft Decision, with a separate report setting out the approach and outcomes of the study in more detail. Telstra engaged Network Strategies to review the WIK-Consult study. The outcomes of the Network Strategies review are set out in a separate report accompanying this submission and summarised at a high level below, alongside Telstra views on other aspects of the WIK-Consult study.

25. It is, however, important to note that the ability to undertake a comprehensive and accurate review

of the WIK-Consult benchmarking study was significantly impacted by the lack of transparency in relation to cost inputs and calculations, as well as the absence of evidence to support many of the assumptions made by WIK-Consult. Telstra does not consider that the information provided by the ACCC or WIK-Consult regarding the approach to international benchmarking is sufficient to justify the proposed primary price term for mobile voice termination. Further, Telstra considers that the ACCC must reassess its acceptance of the mobile voice termination rate calculated by WIK-Consult and that this should be done in broad consultation with industry to ensure accuracy and appropriateness of a primary price term that reflects the efficient cost of providing mobile voice termination in Australia.

3.3.1. Selection criteria for benchmark countries

26. The ACCC requested that benchmarks used by WIK-Consult be based on the following criteria:

a. The benchmark countries should be restricted to those that develop and publish cost models based on a TSLRIC+ (or equivalent) framework. This could include countries which adopt a pure LRIC framework but use models capable of producing TSLRIC+ estimates.

b. The benchmarks to be included must be the outputs of cost models, rather than the regulated rates adopted by regulators (as these may not necessarily align).

27. Based on the above criteria, the benchmark sample selected by WIK-Consult comprises nine

jurisdictions for which the costing model used to develop mobile termination rates is publicly available. Telstra notes that the selection of jurisdictions by WIK-Consult was made purely on the basis of the public availability of a cost model that can produce TSLRIC+ (or equivalent) results. This has resulted in a sample size that is not large enough to minimise the effects of any sampling bias.

28. Further, the selection of benchmark countries has not been based on any criteria which considers whether the country is a reasonable comparator to Australia. That is, unlike similar benchmarking exercises undertaken elsewhere such as New Zealand, the benchmark sample has not been refined to ensure that countries included are comparable to Australia. As set out in Telstra’s submission to the FAD Discussion Paper

6, the aim of selecting a comparable jurisdiction is to

ensure that benchmark costs reflect the conditions under which the MTAS is supplied in Australia. Refining for comparability may include consideration of factors such as urbanisation rates, cost modelling of similar services and population density.

29. In the Draft Decision the ACCC acknowledges that the countries in the benchmark set differ in

characteristics compared to Australia but considers that “...the adjustment process used by WIK-Consult...takes those differences into account when determining the cost of voice termination in

Australia.”7

30. Telstra notes that the ACCC has adopted all of WIK-Consult’s proposed adjustment factors –

Telstra’s position on these adjustments and WIK-Consult’s approach to undertaking the adjustments is set out in Section 3.3.2 below. However, Telstra considers that WIK-Consult’s

6 Ibid.

7 ACCC, Mobile Terminating Access Service – Final access determination, Draft Decision, May 2015, Section 3.2.1



PAGE 10

analysis omits consideration of a range of relevant factors given the benchmark sample selected. Table 1 below compares the benchmark sample across the following characteristics - population density, land area (sq km), urbanisation, number of users per site, number of mobile sites, size of largest mobile network, and coverage % per population.

Table 1 Benchmark countries by population density, land area (km2), urbanisation, users per site, mobile

sites, mobile network size (km2) and coverage %

Country Population

Density8

Land (sq

km) 9

Urbanisation

Rate10

No. Users

per site

(WIK) 11

No. Mobile

sites12

Largest

mobile

network size

(km2) 13

Coverage %

(per

population)

14

Australia 9

7,682,300 89 1,344 Unable to reconcile

[C-i-C begins] [C-i-C ends]

99

Norway 14 365,268 80 430 4,924 411, 221 100

Sweden 24 407,340 86 434 6, 754 333,217 100

Mexico 63 1,943,945 79 2,683 6,659 1,982,059 100

Romania 87 230,020 54 729 11,174 221,852 100

Spain 94 498,800 79 1,013 32,253 516,333 100

Portugal 114 91,590 62 1,392 4,992 92,024 99

Denmark 132 42,430 87 636 2,877 42,258 No data available

UK 265 241,930 82 877 17,204 228,462 100

Netherlands 498 33,720 89 1,444 4,718 33,286 100

31. A number of observations can be made from the comparison in Table 1. Most significantly, there is

a stark lack of comparability across the benchmark sample across a range of factors which influence cost, including population density, land area and mobile network size amongst others. In Telstra’s view, these obvious differences make it particularly difficult to draw any meaningful conclusions regarding the efficient cost of providing mobile voice (and SMS) termination services across the various jurisdictions. Absent adjustments for any of these factors, these significant differences are concerning and call into question the veracity of WIK-Consult’s overarching conclusions. For example, the sheer size of Australia’s mobile networks is incomparable to the

8

Worldbank (2013), Population density –people per square km of land area 9 Worldbank (2013),Land sq km, http://data.worldbank.org/indicator/AG.LND.TOTL.K2 (2014)

10 Worldbank (2013), Urban population, http://data.worldbank.org/indicator/SP.URB.TOTL.IN.ZS

11 WIK-Consult Final Report, , Benchmarks for the Cost of the Mobile Termination Access Service in Australia, 15

April 2015, pg.15 12

WIK-Consult Final Report, Benchmarks for the Cost of the Mobile Termination Access Service in Australia, 15

April 2015, Specific country models as reflected in Appendix. 13

Identifies largest coverage area of mobile network across 2G, 3G and 4G technologies. Replicated from Network Strategies, Benchmarking Mobile Termination Access Service in Australia, Review of the methodology, June 2015, pg. 43 14

Worldbank (2013) Population covered by mobile cellular network, <http://wdi.worldbank.org/table/5.11#>



PAGE 11

European based countries. The closest mobile network in size would be Mexico however Telstra’s network is still approximately 20% larger than Mexico’s total land area. Further:

a. Australia is significantly larger in area than the other benchmark countries. In fact, Australia is almost 400% bigger in terms of land areas than its nearest country comparator. Collectively, Australia’s network would cover almost 65% of the total land (sq km) of WIK-Consult’s benchmark sample (excluding Australia).

b. Whilst Australia’s urbanisation rate is the equal highest in the sample, it is important to consider that there remains a significant segment of the population (i.e. almost 3 million) that reside in regional and remote areas. These areas in Australia span many millions of kilometres. This can be contrasted with the European countries in the sample where mobile deployment across the entire country would be commensurate with Australia’s urban mobile deployment only.

c. Australia remains a significant outlier in terms of low population density. The Netherlands is a significant outlier at the other end of the spectrum, with substantial variation in between.

d. There is significant variation in number of mobile sites across the benchmark sample and a lack of observable relationship between the number of mobile sites and size of network. For example, Mexico’s mobile network which is a significantly bigger network in terms of land area covered has less mobile sites than UK, Spain and Romania. Therefore, drawing any meaningful trend or relationship between number of sites, size of mobile network and population is almost impossible.

e. There appears to be a distinct lack of trend in terms of the number of users per site and population density. For example, the Netherlands which has the highest population density also has the second highest number of users per site. Australia, which has the lowest population density, has only marginally less users per site. The lack of observable relationship is problematic given WIK-Consult use this as part of the basis for dismissing the case for adjustments for population density.

32. Telstra is concerned by the ACCC’s acceptance of a benchmark sample that clearly lacks

comparability across a range of factors. Further, Telstra is most concerned with the omission of population density and consideration of distance/coverage of networks as a relevant adjustment factor in light of the observations made above. Given that WIK-Consult have made no refinement to the benchmark set on the basis of these differences, this means that the adjustment of benchmark costs to reflect differentials in the efficient cost of providing the MTAS between benchmark countries and Australia is even more critical.

3.3.2. Adjustment factors

33. As noted above, given the absence of refinement to the benchmark set, the adjustment of benchmark costs is critical to ensure that they reflect the conditions under which the MTAS is supplied in Australia. WIK-Consult applied adjustments to the derived benchmarks to take into account certain country specific factors that impact the cost of termination services in Australia. These factors are:

a. Currency conversion b. Network technology; share of 2G/3G voice traffic c. WACC d. Network usage e. Geographic terrain f. Spectrum fees

34. Telstra’s positions on the adjustments applied by WIK-Consult are set out below.



PAGE 12

Currency conversion

35. WIK-Consult have converted benchmarks from international cost models to Australian dollars based on an average of the market exchange rate (10 year average) and exchange rate adjusted for purchasing power parity (PPP). This ‘blended’ approach has been adopted by WIK-Consult on the basis that:

a. Some mobile network assets are purchased in international markets, implying that nominal exchange rates are relevant.

b. Some mobile network-related costs are locally sourced, implying that PPP rates are relevant as “Australia is one of the more expensive countries in terms of PPP”.

15

The ACCC considers that this approach is appropriate as applying an average of market exchange rate and PPP-adjusted exchange rate reflects the proportions of these two categories of costs.

36. WIK-Consult notes that the New Zealand Commerce Commission (NZCC) used this approach in its 2011 MTAS proceedings, although Telstra notes that this was done on the advice of WIK-Consult. Outside of that process (and the current FAD Inquiry), Telstra is not aware of any other jurisdictional regulator adopting a hybrid or similar approach for currency conversion in benchmarking exercises. In regulatory benchmarking, currencies are typically converted to the local currency using either current market exchange rates (or an average over a period of time) or PPP rates, but not a combination of market exchange rates and PPP rates.

37. Telstra’s view is that PPP rates alone are the appropriate method of currency conversion for the international benchmarking exercise. The use of PPP as a means of adjusting for different currencies captures:

a. Different wage levels between countries b. Different equipment prices between countries c. Varying capital charges

Using market exchange rates as well as PPP rates to adjust for mobile network input costs effectively leads to double-counting. Further, market exchange rates are subject to volatile capital movements, bearing little or no relation to relative prices or relative inflation rates. As such they cannot adequately adjust for cost differences between countries. The deficiencies of the blended approach used by WIK-Consult are discussed in more detail in the Network Strategies Report accompanying this submission.

Network technology

38. WIK-Consult seeks to account for recognised cost differences in 2G and 3G technology by adjusting for differences in network technology mixes between benchmark countries and Australia. The adjustment aims to take into account differing proportions of voice traffic carried over 2G and 3G technologies and assumes that 6% of voice traffic in Australia is on 2G with the remaining 94% on 3G. The countries in the benchmark set all have a higher proportion of 2G traffic and WIK-Consult therefore seeks to estimate the effect on cost via assumed demand elasticities and resulting changes in traffic volumes.

39. The elasticities used by WIK-Consult are based on a study undertaken for the NZCC which uses

significantly out of date models (2006-2008) and relates only to 2G networks. Given the significant changes in the mobile industry since 2006-2009, it is likely that the elasticities used by WIK-

15

WIK-Consult, Benchmarks for the Cost of the Mobile Termination Access Service in Australia, 15 April 2015, p.24



PAGE 13

Consult do not accurately reflect the effect of changes in demand on current costs. Further, while the cost of 3G technology is lower than 2G, it is not possible to infer elasticity for 3G demand from a 2G study. Nor is it possible to infer the relativity between a 2G and 3G demand elasticity.

40. Network Strategies reviewed the adjustment for network technology made by WIK-Consult and

found that the changes in demand associated with the adjustments are far greater than that used for its elasticity estimate. Specifically, WIK-Consult reduced 2G traffic volume between 81% and 91% while increasing 3G traffic volume between 38% and 194%. Elasticity is normally not constant along the cost curve and it is therefore invalid to apply WIK-Consult’s elasticity estimate in these cases. Further, the very low volume of 2G traffic assumed for Australia implies that the 2G network is driven purely by coverage requirements rather than capacity. The costs of 2G networks are therefore largely fixed (rather than volume-dependent). WIK-Consult’s elasticity estimate would also not apply in this instance.

41. Based on the above, Telstra considers that there is no evidence to support WIK-Consult’s elasticity

assumption and the associated adjustment for network technology. Although the underlying methodology is reasonable, the absence of a robust elasticity assumption means that it is largely abstract.

WACC

42. Telstra agree with the ACCC that it is important to make adjustments for the WACC to ensure that the difference in cost of financing capital expenditure in Australia is taken into account when setting regulated MTAS rates for both voice and SMS.

43. However, Telstra has significant concerns with the approach used by WIK-Consult in adjusting the benchmark results to account for differences in the WACC, as well as the use of the Fixed Line Services Draft Decision WACC.

44. With respect to the choice of the WACC from the Fixed Line Services Draft Decision, although

Telstra agrees that it is sensible and reasonable to apply the results of WACC considerations from that process to the MTAS FAD, as set out in Telstra’s submission in response to the ACCC’s Fixed Line Services Draft Decision (May 1 2015), the overall WACC values, and the parameter choices that contribute to the WACC set out in the Draft Decision, are not reasonable. Telstra’s concerns are summarised below (from paragraph 48) and set out in detail in Telstra’s May 2015

submission.16

45. Further, the manner in which WIK-Consult have adjusted the benchmark values to account for

differences in the WACC values is opaque, relies on dated modelling results and avoids the most straightforward and transparent approach. Specifically:

a. WIK-Consult apply adjustment factors to account for differences in the WACC among

benchmark countries and Australia that are derived from a near ten year old model developed by WIK-Consult for the New Zealand Commerce Commission. As noted by Network Strategies, “The models used in WIK’s study are now almost ten years old, and relate solely to 2G technology. Given changes in the mix of technology – in particular with the effect of lower costs associated with 3G – it would be inappropriate

to apply the resultant elasticities to current models.”17

Further, “WIK’s sensitivity

analysis for the WACC was based solely on the impact of increasing the WACC from 10% to 15%. As WIK has noted, the value of the elasticity is not constant along the

16 Telstra, Public inquiry into final access determinations for fixed line services—primary prices, Response to

Draft Decision, 1 May 2015 17

Network Strategies, Final Report, Benchmarking Mobile Termination Service In Australia- Review of the

Methodology, Report 35012, 5 June 2015 pg. A42



PAGE 14

cost curve. WIK’s application of this elasticity to the current situation in which the various WACC values – ranging from 6.29% to 12.95% – are reduced to 5.43% may

therefore be misleading.”18

b. Overall, Network Strategies consider that the use of such a dated model (based on 2G technology) and the use of a constant elasticity likely mean that “WIK’s use of the

elasticity estimate for the WACC adjustment is invalid.”19

46. More importantly the approach used by WIK-Consult is needlessly opaque. WIK-Consult have

access to the full public models used in the benchmark sample. Therefore, WIK-Consult can simply apply the relevant WACC to each model in order to directly observe the consequent change in MTAS results, without having to apply a secondary-sourced adjustment factor. There is no reason given as to why this approach is not used by WIK-Consult in making the necessary adjustments to the benchmark results to account for differences in the WACC.

47. Beyond these concerns regarding the approach used by WIK-Consult to make the necessary WACC adjustments, Telstra reiterates the concerns expressed in our May submission to the Fixed Services FAD Inquiry that the WACC set out in the Draft Decision of that process is not reasonable and is based on incorrect parameter estimates for key model variables. In its response to the Fixed Line Draft Decision submitted in May 2015, Telstra provided detailed commentary regarding the ACCC’s assessment of the WACC and has highlighted significant errors in its computation. Telstra reiterates these concerns below, and considers that setting a WACC of 5.43% will underestimate the cost of providing services in Australia relative to other countries in the rest of the benchmark sample.

The WACC underestimates the cost of financing capital expenditure in Australia

48. It is reasonable that the ACCC use the WACC determined for Telstra as the benchmark efficient operator given that Telstra is the largest operator, with the best credit ratio and thus, likely lowest debt issuance costs (which are based on the profile of the integrated company). In addition, as two out of three of the Australian MNO are highly integrated players and provide services across both fixed and mobile, Telstra consider that the consistency across Determinations is entirely appropriate. Finally, Telstra do not consider that the risk relative to financing of mobiles is inherently different to that of fixed infrastructure. Whilst both technologies are at different stages of their product lifecycle and evolution (with mobiles still evolving and fixed at an end stage), Telstra consider the risk profile may be commensurate. Whilst mobile networks continue to evolve at rapid pace, return is uncertain as the technology can quickly be superseded. Whilst for fixed networks, overbuild with next generation broadband networks means that any future capital investment is less than certain. In the absence of better information, Telstra concurs with the ACCC’s principled approach to the consistency of WACC application cross both fixed and mobile services.

49. In Telstra’s response to the ACCC’s Fixed Line Services Draft Decision, Telstra supplied ample

evidence demonstrating that the ACCC’s draft determination of the WACC was too low. As Telstra stated in its response to the Draft Decision, the WACC “... is also the lowest WACC set by any Australian regulator in any decision that Telstra is aware of over the past two years. WACC values determined by other regulators over this period have ranged from 5.93% to 10.42%, with an average of 7.16%. The majority of decisions over this period have delivered an overall WACC in

the range of 6.5% - 7.5%”20

and “[t]he WACC adopted in the Draft Decision is also significantly

lower than any recent estimate of Telstra’s cost of capital by independent market practitioners,

such as brokers or analysts”21

and “[t]he ACCC’s WACC is more than 1% lower than the lowest

18

Op.Cit 19

Op.Cit 20 Telstra, Public inquiry into final access determinations for fixed line services—primary prices, Response to

Draft Decision, 1 May 2015, pg. 144 21

Op.Cit, pg. 27



PAGE 15

estimate from recent broker reports and nearly 3% lower than the median estimate from these

reports (8.2%).”22

50. We address some specific aspects of the WACC below. Equity Beta

51. In the MTAS Draft Decision, the ACCC highlight that the equity beta is of particular relevance. Specifically, the ACCC note that: “The equity beta of 0.7 adopted in the fixed line services FAD draft decision is consistent with the outcome of benchmarking equity betas for comparable telecommunications service providers across OECD countries. These comparable telecommunications services providers provide a range of telecommunications services, such as fixed line and mobile services. Therefore, the ACCC expressed the view in the fixed-line services FADs draft decision that the equity beta of 0.7 is likely to be higher than that of an operator who provides fixed-line services alone. The ACCC considers that this equity beta is appropriate for a hypothetical efficient MNO in Australia as two out of the three MNOs are integrated operators.”

23

52. Specifically, Telstra note that the equity beta should be increased to at least 0.8, in order to

properly compensate for risk exposure. In Telstra’s October 2014 submission in relation to Fixed Services, Telstra highlighted a number of differences between Telstra and other regulated businesses which mean that it is likely to be more exposed to systematic risk relative to other firms, including differences in the nature of services supplied, and particular, higher income elasticity of demand for telecommunications services, and differences in the form of regulation applied to Telstra and its infrastructure peers.

24

Debt Risk Premium

53. In relation to the debt risk premium (DRP), the ACCC has relied on a new and untested data source without considering whether the estimate produced by this source is reasonable, resulting in a DRP that is lower than any recent decision ACCC decision and significantly below Telstra’s efficient cost of debt financing – which has been recently tested and identified by Telstra placing a long term bond in international markets. On 30 March 2015, Telstra issued a 10-year bond in the US market. [C-i-C begins] [C-i-C ends]. After swaps to convert the US$1 billion issue into Australian dollars, the semi-annual coupon rate attached to this bond issue was 4.27%, implying a DRP of 1.94%.

25 This is more than double the value set out by the ACCC in the Draft Decision.

Market Risk Premium

54. The ACCC's estimate of the market risk premium (MRP) does not reflect prevailing market conditions. The ACCC's estimate is out of step with current empirical evidence and the views of other regulators, particularly the Australian Energy Regulator (AER), and tends to significantly understate the MRP. Based on detailed analysis of the AER’s most recent decisions, Telstra considers that it would be unreasonable for the ACCC to maintain an MRP of 6% in determining prices for the declared fixed line services. The evidence supports an MRP of at least 6.5%.

26

22

Ibid. 23

ACCC, Mobile Terminating Access Service, Final access determination, Draft Decision, May 2015, section 3.2.1 24

Telstra, Public inquiry into final access determinations for fixed line services—primary prices, Response to

Discussion Paper, October 2014 25

Telstra, Public inquiry into final access determinations for fixed line services—primary prices, Response to Draft

Decision, 1 May 2015, p.150 26

Telstra, Public inquiry into final access determinations for fixed line services—primary prices, Response to Draft

Decision, 1 May 2015, p.154



PAGE 16

Network usage, network size and population density

55. The ACCC has made a draft decision to adjust the benchmarks for network usage (per mobile cell site) but to not include any adjustments for population density or network size across the benchmark sample. This is based on analysis by WIK-Consult, which has compared the network usage of operators in Australia (GB per site) with the other comparator countries and concludes that Australia’s network usage in terms of GB per site is significantly greater than the other countries. WIK- Consult then makes adjustments based on an estimated elasticity of reaction of cost to the degree of network usage. The adjustment for network usage resulted in an average decrease of the benchmark of 0.115 AU cents.

56. In their report, WIK-Consult dismiss the generally accepted view that population density is a key

driver of cost on the basis of a limited analysis. WIK-Consult compares population per mobile site and volume of traffic per site (GB) across the benchmark sample. Although WIK-Consult acknowledge that population density can be a significant driver of the cost of deploying and maintaining mobile networks, they dismiss this as a significant cost contributor on the basis that Australian networks have on average, a comparable number of users per site and a considerably higher volume of data traffic (in GB) per site when compared with the benchmark sample. Based on WIK Consult’s analysis, they conclude that average per unit cost of traffic is comparable to the benchmark countries and no adjustment for population density is therefore necessary.

57. The conclusions reached by WIK-Consult are out of step with international precedent, as well as decisions previously made by the ACCC and analysis undertaken by WIK-Consult. The ACCC previously considered population density an important factor in network costs.

“The Commission believes there are many factors that influence the cost of the MTAS in different jurisdictions, including: geographic terrain, population density, Network usage and scale, land and labour costs in different jurisdictions, spectrum allocations, the extent to which mobile operators are integrated fixed and mobile network operators, network purchasing power, cost of capital in different jurisdictions and the mobile network technology deployed in different countries.

27(emphasis added)

58. It is hard to conceive why WIK-Consult has deviated from accepted norms, including network

scope and population density (and the interrelationship between these factors and the impact on the cost of coverage) in developing and applying adjustment factors for MTAS. For example, the NZCC concluded, based on advice from WIK-Consult that:

“...population density, taken together with urbanisation rates, influences the proportion of coverage driven network elements in a mobile network. The Commission has taken this factor into account in selecting the price point, based on the results of its benchmarking”

28

The NZCC also cited the importance of population density in setting prices. Specifically, they stated: “This implies that Australian networks will be less efficient than networks in New Zealand because there will be proportionally more coverage-driven network elements, increasing average costs for a given level of coverage. Having a greater proportion of coverage-driven network elements in Australia will also result in average link distances being higher in Australia, resulting in higher costs.”

29

59. The fact that WIK-Consult dismiss population density (and related network cost factors such as

geographic coverage and coverage per head of population) reflects the overly-simplified analysis undertaken in the benchmarking report and the adoption of unsupported assumptions regarding

27

ACCC, Mobile Services Review: Mobile Terminating Access Service, Final Determination, June 2004, p. 214. 28

NZCC, Final Decision, May 2011, pg 54 29

NZCC (2011), p.79



PAGE 17

the relationship between network usage (on average) and the costs of deploying networks in different jurisdictions with wildly different coverage requirements and population distributions.

60. Network costs are highly divergent depending on where the networks are deployed. The cost of

deploying mobile network infrastructure in urban environments is considerably less expensive than deploying infrastructure in rural and remote areas. This reflects higher capital costs (including the need for more extensive transmission infrastructure) to support network deployment in regional and remote areas, as well as higher operating costs – reflecting increased travel time and penalty rates to support remote network infrastructure.

61. Put simply, by only looking at aggregated and average coverage per cell site and usage per cell

site, WIK-Consult fail to account for the significant divergence in costs for deploying geographically very large networks (as is the case in Australia) compared to relatively small networks in European countries used within the benchmark set. Further, and related to this point, WIK-Consult have failed to account for the variation in cell site costs between low-cost, high density urban areas and higher-cost, low density regional and remote areas. Given the relatively greater presence of higher cost, low density coverage sites in Australia it is necessary to explicitly understand and take into account the impact of different cell site density and network coverage in benchmarking network costs.

62. Compared to the networks in WIK-Consults benchmark sample, Telstra’s network is significantly

larger and is likely to contain far more sites in outer regional/remote/very remote areas than networks in the benchmark sample. This means the distribution of sites will differ significantly as will the type of network infrastructure required in Australia, including kilometres of fibre optic cable, number of transmission systems and optical repeaters. This network infrastructure will be significantly different and more costly than what is required to dimension geographically smaller networks in other jurisdictions. However none of these distinct attributes are considered in WIK Consult’s simplified analysis.

63. Population density measures population over land area (in sq km). However, in dismissing

population density as a relevant measure, the ACCC and WIK-Consult make no allowance for distance (in terms of land area and size of mobile network and the implications this has for distance between cell sites in Australia). In dismissing population density as a relevant network cost factor, WIK-Consult consider a number of alternate information sources, such as number of mobile users, number of cell sites and the volume of traffic to ultimately conclude that Australia’s outlying population density or significant mobile network size is not a relevant cost consideration. In Telstra’s view, this is a significant flaw.

64. Network Strategies note the deficiencies in WIK-Consult’s analysis and find that: “In its use of network usage expressed as traffic per site, WIK attempts to adjust for differences in network scale, however this only partially addresses the coverage issue. Mobile networks with larger coverage areas will have more base stations, and – especially for rural areas – are likely to have more backhaul, expressed in terms of total distance, so that the average backhaul distance per site is likely to be greater. This will clearly have an effect on costs for fibre or leased line backhaul.”

30

Network Strategies’ views align with Telstra’s own network cost data. Telstra’s own analysis is discussed further below.

65. As part of their analysis, WIK-Consult calculated the average number of mobile users per site across the benchmark sample and found that the number of users per site for Australia is comparable with the benchmark sample. Specifically, WIK-Consult state:

30

Network Strategies, Benchmarking Mobile Termination Access Services in Australia, June 2015, p.A43



PAGE 18

“As far as the number of users served on average from one site is concerned, the number for Australia is lower than that of only three of the benchmark models and is higher than that of the other six. What does this mean? Even if there were relatively many coverage-driven cells in Australia with few users being served in them, this becomes irrelevant, given the large number of users served in other sites. It is clear that the more users are served per site, the lower is the cost per unit of service, since the fixed cost of a site is spread over more and more users. Therefore, it follows already from this comparison that the cost of termination need not necessarily be higher in Australia due to low population density and the resulting prevalence of coverage driven cells.”

31 (emphasis added)

This is at best speculation and is not supported by any evidence or analysis. On the other hand there are a number of reasons to conclude that the cost of supplying voice services on Australian networks will be higher than those of other countries in the benchmark sample.

66. First, WIK-Consult have failed to assess the impact of coverage and network scope as a cost driver and a source of difference between benchmark networks, which are generally only a fraction of the size and coverage of Telstra’s network (and other Australian mobile networks). [C-i-C begins] [C-i-C ends]

67. Second, Australia’s mobile networks are considerably larger than the networks being benchmarked

against. The following table is reproduced from Network Strategies’ report and shows that other than Mexico, the benchmark sample country networks are significantly smaller in geographic scope than Telstra’s network and other Australian networks.

31

WIK-Consult, Final Report: Benchmarks for the cost of the Mobile Termination Access Service in Australia, April

2015, p.14



PAGE 19

Table 2 Coverage areas of modelled networks

Country Coverage area (km2)

2G 3G HSPA 4G Total

Denmark 42,258 42,258

Mexico 1,982,059

Netherlands 33,286 24,051

Norway 411,221 202,135

Portugal (2012)1

92,024

Portugal (2015)1

92,024

Romania 221,852 152,079

Spain 499,145 510,237 516,333

Sweden2

333,217 272,042 163,863 2,552

United Kingdom (2014) 228,462 201,158 89,435

United Kingdom (2015) 228,462 201,158 89,435

1 No breakdown of coverage by technology was available for Mexico or Portugal

2 Coverage for Generic integrated operator scenario

Source: regulators’ cost models

68. WIK-Consult omit the consideration of network size in their benchmark analysis. As Network Strategies note:

“There are network characteristics other than just the technology mix that may result in cost differences between the various sample countries and Australia. While WIK examine differences in network usage, it does not consider the scope, or coverage of the network.”

32

This is a significant omission.

69. Irrespective of the number of users per site, undertaking analysis and drawing conclusions without

reference to geographic coverage understates the true cost of building and maintaining networks. Greater geographic coverage (coupled with Australia’s low population) means that even though the average number of users per network cell site may be similar in Australia and the benchmark sample countries, the scope and scale of Australian networks and the distribution of population coverage is dramatically different. The combination of these factors mean that the costs of deploying network infrastructure on a per user basis will very likely be (significantly) higher in

Australia than for benchmark sample. Error! Reference source not found. presents some of

Telstra’s own analysis on the costs to deploy in different areas of Australia. [C-i-C begins] [C-i-C ends]

32

Network Strategies, Op. Cit., pg. A43



PAGE 20

Traffic volume per site is a secondary consideration when benchmarking voice termination

70. The primary function of mobile networks is to supply voice services. In particular the costs incurred in providing coverage to remote areas are driven by the requirement to provide contiguous network coverage to avoid call dropouts and coverage blackspots. Although the evolution of mobile networks over time has seen users complement their voice calling with data usage, the resulting allocation of costs is far more complex than the simple comparison undertaken by WIK-Consult would suggest. The lack of rigorous analysis is problematic and leads to conclusions that bear no resemblance to how mobile networks are dimensioned and costs apportioned and have led WIK-Consult to draw incorrect conclusions and adjustments based on network usage data.

71. Network Strategies conclude:

“We do not agree with WIK’s claim that it has demonstrated via its network usage measure (total traffic in GB per site) that the cost of voice termination should be lower in Australia than in the benchmark countries due to the higher traffic volumes. Traffic per site will only provide a partial –and thus incomplete-indication of cost differences due to network scale. Consequently we find that WIK’s adjustment is inappropriate and may be misleading.”

33

72. Further Network Strategies note the lack of replicabilty of results, particularly for Denmark, Portugal

and Sweden where more recent versions of the model are available. Using an updated model result in more traffic per site (with the exception of the UK). This clearly calls into question, WIK-Consult’s conclusions regarding the degree of discrepancy of traffic volume between comparator countries. Given this, Telstra therefore caution the ACCC to consider the emphasis it places on this analysis.

73. Despite these deficiencies, Telstra note that WIK’s analysis of volume of traffic per site in GB concludes that Australia has a significantly higher volume of traffic than the benchmark countries. WIK-Consult, in error then goes on to conclude that Australia’s low population density is irrelevant in determining voice termination costs. The ACCC state:

“WIK-Consult noted Australia has a very high usage per cell site compared with the networks usage evident in the benchmark counties. It concluded that given the high network usage cell site in Australia, low population density does not actually mean that the average cost of traffic in Australia is higher than in other countries.”

34 (emphasis added)

74. Telstra disagrees with this conclusion and considers that:

a. First, for the reasons outlined earlier, there are significant differences between Australia and the remaining benchmark countries which account for significant differences in cost of mobile network deployment.

b. WIK-Consult’s analysis over emphasises the impact increasing data usage has on the cost of deploying and maintaining mobile networks. In facilitating voice over the mobile network, Telstra incurs additional cost relative to data as it must ensure that the voice call meets certain required standards. In contrast, data transmission is always simply a best effort service. Simply, the mobile network can always transmit a data packet later if necessary, but voice requires more accuracy on the first transmit to be intelligible and useful to end consumers. [C-i-C begins] [C-i-C ends]. Therefore, it is not legitimate to assume (simply) that, as average data traffic is high, more and more costs should be apportioned to data, in place of voice. Voice can be

33

Network Strategies, Op. Cit, p.A45 34

ACCC, Draft Decision, Section 3.2.2



PAGE 21

characterised as a superior function over the network and as such, voice and data are not of equal value on the network. Simply looking at the quantum of traffic is therefore misleading.

c. The volume of traffic is not equal across all areas of Australia. For example, Telstra’s own analysis [C-i-C begins] [C-i-C ends] Even if WIK-Consult’s analysis was valid (which Telstra disputes), WIK-Consult does not consider the traffic profile across Australian sites. [C-i-C begins] [C-i-C ends]

d. An analysis of Telstra’s network architecture confirms [C-i-C begins] [C-i-C ends]. It is unclear if, and how, WIK-Consult have considered this in their simplified analysis.

e. Even if WIK-Consult’s simplified interpretation were to be correct, it omits a significant aspect. If Australian networks carry on average 350%

35 more traffic per site than the

other benchmark comparators, WIK-Consult’s analysis does not capture the cost of delivering this additional traffic, i.e. it is likely that Australian networks would have incurred additional capital expenditure relative to the other benchmark countries to facilitate the additional traffic which would need to be recouped.

f. WIK-Consult’s analysis disregards the inherent bias in examining traffic volumes from the benchmark models. The models are built with the purpose of ascertaining efficient mobile termination costs and are therefore likely under dimensioned for data. In short, this is not a like-for-like comparison.

g. Finally, it is unclear how and from where the data relating to data usage per GB for Australia has been derived. It is highly unlikely based on data we have seen that Telstra’s volume of data on mobile networks surpasses the other sample countries to the extent displayed. As has been discussed elsewhere in this submission, Telstra remains concerned at the lack of detail and opaqueness of the WIK-Consult analysis.

75. In Telstra’s view, WIK-Consult have incorrectly emphasised the importance of data traffic to the issue of cost of deployment and maintenance and incorrectly relied on this anomaly to justify a lack of adjustment based on density and to produce an adjustment based on network traffic only. The general principle underpinning the economics of mobile networks is that for a given level of traffic, the greater the coverage area means higher costs. All of these reasons, combined, show that WIK-Consult’s analysis is too simplified and is therefore not appropriate.

Geographic terrain

76. Recognising that the geographic features of Australia are likely to be different from those of the benchmark countries, the ACCC considers it appropriate to make adjustments to the extent that such features impact on the cost of voice termination services in Australia.

77. Telstra notes that, while it is widely acknowledged that geographic terrain will have an impact on the cost of a mobile network, it is difficult to quantify the relative differences in terrain between countries. Network Strategies have advised Telstra that there is no international standard for terrain classification and, as such, any attempts at classification and adjustment are highly subjective and subject to considerable uncertainty.

78. The WIK-Consult approach to making adjustments for geographic terrain is based on classifying

benchmark countries based on comparing ‘mountainous territory’ to Australia. The classification is based on a visual inspection of topographic maps. Although unclear, this inspection presumably encompasses entire countries rather than mobile coverage areas which is likely to introduce a further degree of subjectivity and error. This is because in some jurisdictions mobile coverage extends over almost the entire country, while others still have significant areas without coverage. Telstra considers that any classification relating to geographic terrain should therefore be restricted to mobile coverage areas.

35

Average volume of traffic per site in GB (for all benchmark countries excluding Australia)



PAGE 22

79. While mountainous territory does have an impact, the propagation of radio signals is also affected by other topographic features, such as buildings, vegetation and weather conditions. These can significantly attenuate radio signals, requiring sophisticated planning tools to estimate path loss and signal strength at various locations. Telstra therefore considers that any cost adjustments for geographic terrain should take account of these additional topographic features rather than being limited to mountainous territory.

80. That said, while the WIK-Consult approach is incomplete in terms of the topographic features

adjusted for, Telstra considers that that small scale of the adjustment applied by WIK-Consult and the inherent uncertainty associated with such an adjustment suggests that it would be preferable for the ACCC not to include any adjustment for geographic terrain.

Spectrum fees

81. The ACCC has included an adjustment for the differences in the cost of spectrum that MNOs have to incur to provide mobile services in Australia. This is one of the larger adjustments made to the benchmarks. WIK-Consult has applied the following approach to adjust for spectrum costs:

a. Set cost components for spectrum in benchmark models to zero; b. Derive a per-unit mark-up to account for spectrum fees from ACCC-provided

information on the amount paid by MNOs for obtaining spectrum to provide mobile services and the term of spectrum licences in Australia;

c. Add a per-unit mark-up to the benchmarks for voice termination with local spectrum fees removed.

As part of this process, WIK-Consult added a 2% opex over the annualised spectrum cost to account for costs incurred by MNOs in using the spectrum.

82. The adjustment applied by WIK-Consult was reviewed by Network Strategies who found that, although the adjustment accounts for differences in spectrum fees, it fails to take account of the implications of those differences. That is, there are a number of reasons for differences in spectrum costs (such as the quantity and mix of spectrum bands) and variations between benchmark countries are likely to have an influence on network costs which also need to be taken into account.

83. By simply removing network costs – as WIK-Consult has done – there has been no consideration of the influence of the characteristics of the various operators’ spectrum holdings on network costs overall. For example, depending on the type and quantity of spectrum in question, an operator who has incurred relatively high spectrum fees may have then been in a position to deploy a lower cost network than an operator who has incurred relatively lower spectrum fees (but incurs higher network costs). In failing to take the broader impact of the spectrum fees paid by the benchmark countries into account, the WIK-Consult adjustment for spectrum fees represents only a partial adjustment for the effect of spectrum fees on mobile termination costs.

Review of adjustment process

84. Network Strategies reviewed the adjustment for spectrum fees applied by WIK-Consult and found a number of deficiencies, including:

a. A difference in value of spectrum fees for Norway in the calculations for 2G and 3G

LRIC+++ of 4% (2G) and 6% (3G). b. A difference in the value of spectrum fees for Portugal of 0.4% due to WIK-Consult

excluding only the upfront cost for spectrum but including yearly spectrum fees. c. The assumption that a hypothetical efficient operator incurs spectrum fees across a

number of bands does not reflect spectrum holdings by Australian MNOs.



PAGE 23

d. The averaging of estimated fees across the three Australian MNOs has not been correctly applied by not taking into account that all MNOs have not acquired spectrum in all bands.

e. The spectrum costs assumed by WIK-Consult may represent an inefficient allocation of spectrum.

f. Use of a simple annuity rather than tilted annuity. 85. These issues, and others, are set out in more detail in the Network Strategies report. Further,

Network Strategies considered that WIK-Consult’s spectrum fee adjustment could be improved by:

a. Making an explicit assumption regarding the spectrum holdings of the hypothetical efficient operator, thus providing more transparency for ensuring that it is an appropriate proxy for an Australian MNO.

b. Using tilted annuities in the annualisation calculation. Voice-over-LTE (VoLTE)

86. In the Draft Decision, the ACCC considers it highly likely that all three MNOs will launch VoLTE services during the next FAD with a resulting impact on the cost of termination services given the efficiency gains associated with the use of this technology. However, the ACCC has made a draft decision not to adjust the mobile voice termination rate to take account of VoLTE as it considers that it does not currently have reliable information on the potential take-up of VoLTE services to be able to do so.

87. Telstra agrees with the ACCC draft decision not to adjust the mobile voice termination rate for VoLTE. Although announcements have been made regarding the potential launch of VoLTE in Australia, Telstra considers that the timing of deployment is not yet certain due to the need to develop industry standards and ensure the availability of compatible handsets. The latter of which is largely outside of the MNOs control.

88. In any event, even where VoLTE services are launched during the next FAD, Telstra considers it

unlikely that deployment will occur to the extent that it impacts on the efficient costs of providing mobile voice termination. As Telstra has previously submitted, while 4G technology (including VoLTE) may have the potential to lower the cost of supplying the MTAS in Australia in the future, it is too early to estimate what the actual impact (if any) may be. In the short term it is likely that VoLTE will be less technically efficient that 3G voice, although this will improve over time as technical enhancements are introduced. It is therefore possible that the unit costs of MTAS provision will actually rise in the initial stages of deployment not least because MNOs will need to support the supply of the MTAS over both 3G and 4G/LTE networks.

89. At this point in time, any adjustments to mobile voice termination rates to take account of VoLTE

would be speculative and risk regulatory error. Further, the use of forecast VoLTE traffic shares in the cost model of an international jurisdiction which differs significantly from Australia in terms of structure and network to estimate the impact of VoLTE, as used by WIK-Consult, would be inappropriate.

90. While the ACCC has made a draft decision not to reflect the (potential) deployment of VoLTE, the

Draft Decision notes that this position may be reviewed during the term of the FAD if there is sufficient evidence that the mobile voice termination rate no longer reflects the efficient cost of mobile voice termination in Australia. Telstra considers that the ACCC should exercise caution in considering whether to include the impact of VoLTE, either now or while the FAD is underway. Rather, given that the impact of VoLTE during the next FAD is likely to be insignificant (or result in a short-term increase in efficient costs) and the industry as a whole will be in a state of significant



PAGE 24

transition, Telstra’s position is that the costs associated with revisiting this position36

are likely to outweigh any benefits.

3.4. WIK-Consult international benchmarking approach

91. As part of the review of the approach taken by WIK-Consult, Network Strategies identified a number of issues with the application of international benchmarking. Some of these are outlined in the relevant preceding sections. This section provides a high level summary of those issues with more detail provided in the accompanying Network Strategies report. It should be noted that the lack of transparency associated with the WIK-Consult approach and calculations have limited the ability to undertake a comprehensive review.

92. Network Strategies found that the benchmarks do not include the latest available information.

Specifically:

a. Denmark – WIK-Consult have used the 2012 version of the Danish regulator’s cost model where more recent versions are available. The more recent model, which was used to set prices for 2015, has very different demand forecasts which will have implications for the allocation of costs to voice termination.

b. Portugal – the Portuguese regulator has recently released a preliminary 2015 model for consultation which includes updated forecasts which, as with Denmark, have implications for the allocation of costs to voice termination. Although Telstra recognises that there may be reservations with using a preliminary cost model, the implications of this should be taken into account.

c. United Kingdom – the mobile network cost model for the UK has recently been updated and published by Ofcom. However, WIK-Consult has based benchmark values on an earlier version. The 2015 version of the UK model includes up-dated demand projections and is more reflective of the current state of the market.

93. Network Strategies also considers that there is an error with the way that WIK-Consult have treated

common costs. Specifically:

a. WIK-Consult have assumed a mark-up of 10% for non-network common costs although this does not align with the values included in the benchmark sample cost models.

b. Telstra considers that non-network common costs should be removed from the benchmark rates with a value then added that is appropriate for an efficient Australian MNO.

3.5. Draft mobile voice termination rates for 2016-2019

94. The Draft Decision includes a primary price term of 1.61cpm for mobile voice termination based on the international benchmarking study conducted by WIK-Consult. As noted above, Telstra considers that the WIK-Consult study is flawed and, in its current state, does not provide the level of evidence required by the ACCC to justify the introduction of a new primary price term for mobile voice termination.

95. Telstra also considers that the lack of transparency regarding the approach taken by WIK-Consult, along with the absence of supporting evidence for a number of the assumptions made by WIK-Consult, mean that it is difficult for any meaningful consultation on the proposed primary price term for mobile voice termination. Despite publicly available cost models being employed, many of the calculations made by WIK-Consult were unable to be replicated by Telstra or its external experts.

36

Such costs include those associated with re-opening the FAD, regulatory processes and the provision of

information by MNOs on VoLTE traffic and forecast demand.



PAGE 25

Further a simple comparison of the proposed primary price term for mobile voice termination in Australia compared to other jurisdictions suggests that the 1.61cpm rate cannot be supported. Figure 2 below shows that the proposed rate is the lowest TSLRIC+ voice termination rate amongst a sample of 30 international jurisdictions, as well as being significantly below many jurisdictions which employ a Pure LRIC methodology.

Figure 2 International mobile voice termination rates

96. The implication that the TSLRIC+ based estimate of the efficient cost of providing mobile voice termination services in Australia is only marginally above the Pure LRIC based estimate for the UK and below that for the Netherlands supports Telstra’s view that the approach taken by WIK-Consult is flawed. The nature of the mobile network in Australia, in particular network topography, make it very unlikely that the costs of mobile termination are on par with the UK or indeed many other densely populated European countries. As a result Telstra is concerned that the benchmark rate calculated by WIK-Consult significantly understates the efficient cost of providing mobile termination in Australia.

0 2 4 6 8 10 12

Romania

Malta

Denmark

Norway

France

Sweden

Austria

United Kingdom

WIK benchmark

Netherlands

Belgium

Italy

Hungary

Spain

Cyprus

Croatia

Slovenia

Czech Republic

Mexico

Portugal

Germany

Turkey

Poland

Slovakia

Kenya

Bulgaria

Malaysia

South Africa

Nigeria

Tanzania

Indonesia

Voice termination rate (AUD cents)

Pure LRIC

LRIC+



PAGE 26

04 SMS TERMINATION SERVICES 97. The Draft Decision sets out the ACCC’s position on the approach to pricing SMS termination

services. Specifically that:

a. The ACCC should set a regulated price for SMS termination.

b. The SMS termination price is based on a two-part approach incorporating a conversion factor and SMS-specific cost.

c. A rate of 0.03 cents per SMS will apply from 1 January 2016.

Telstra’s position on the above is set out below.

4.1. A regulated price for SMS termination

98. The ACCC has reached a draft decision to set a regulated price for SMS termination as part of the

current FAD Inquiry. This decision has been made on the basis that it would be inappropriate not to include a regulated price for SMS termination given that the ACCC has previously concluded that the declaration of SMS termination is necessary to promote the LTIE and facilitate cost-based pricing for SMS termination. In considering the declaration of SMS termination, the ACCC position was that commercial SMS termination rates have been well above costs for many years and this may have constrained the ability of some MNOs to offer more competitive retail SMS packages.

99. Telstra continues to consider that there is no requirement to set a regulated price for SMS

termination. Since the ACCC made its Final Declaration Decision in June 2014, the mobile sector in Australia has continued to develop to the extent that the basis of the ACCC’s decision in relation to the declaration of SMS is uncertain.

100. Specifically, one of the drivers for the declaration decision was that the ACCC concluded that

there are no effective substitutes for SMS termination services. The ACCC considered that OTT messaging applications were not an effective substitute as these messaging applications require a smartphone, which only a portion of Australians used, and a data connection.

37 The smartphone

penetration rate relied on by the ACCC was an ACMA figure showing that 49 per cent of Australians used a smartphone. This figure is now significantly out of date.

101. At the end of 2014, estimates of smartphone penetration ranged from 75 to 90 per cent

penetration,38

with Australia widely acknowledged to have one of the highest mobile and smartphone penetration rates worldwide.

39 The ongoing increase in smartphone penetration, as

well as improved 4G coverage and the availability of wifi, suggests that higher numbers of Australian mobile consumers are able to access OTT messaging platforms as an alternative to SMS. This is supported by recent Ofcom research which showed a “surge” in instant messaging on mobile social media platforms, such as Facebook Messenger and Whatsapp.

40

37

ACCC, Domestic Mobile Terminating Access Service Declaration Inquiry – ACCC’s Final Decision, June 2014,

p.33 38

See for example: http://www.whatech.com/market-research-reports/press-release/telecommunications/58149-

new-report-australia-mobile-communications-statistics-and-forecasts 39

http://www.smartcompany.com.au/technology/43413-australia-has-one-of-the-world-s-highest-rates-for-tablet-and-

smartphone-ownership-pc-use.html 40

Ofcom ‘Media Uses and Attitudes 2015’ as reported by Telecom.com



PAGE 27

102. Mobile service providers continue to respond to the increasing popularity of OTT messaging application by adapting their mobile plans to, amongst other things, more widely include ‘free’ SMS and increase data allowances. Telstra has consistently submitted that the declaration of SMS termination will have a limited impact on the downstream retail market as SMS traffic is relatively balanced between MNOs and the majority of mobile phone plans include unlimited SMS. The increased competitiveness of the mobile sector – generally and in response to the evolving OTT market - suggests that there is no need for the ACCC to set a regulated price for SMS. This is particularly as the process for setting a regulated price is, as detailed below, subjective and subject to regulatory error.

103. Where the ACCC is minded to set a regulated price for SMS termination, Telstra considers

that A2P SMS should be excluded from the application of primary price terms. This is because A2P termination services are distinct from traditional SMS termination services in a number of ways. The provision of bulk SMS is an emerging or nascent service compared to traditional SMS and raises issues of potential congestion and the management of risks around spam. Importantly, they are provided to bulk SMS providers or aggregators as an end-to-end service rather than a standalone termination service (as provided between MNOs).

104. The distinct nature of A2P termination services means that, much like MMS, they should be

treated as a separate market from traditional SMS. Unlike traditional SMS which is relatively stagnant (due in large part to the popularity of OTT applications), the market for A2P messaging continues to grow and evolve. Applications for use range from providing reminders and time-critical alerts to banking services and payments, to mobile marketing, customer service notifications and authentication processes. On a global basis, forecasts predict that A2P SMS will be worth almost $60 billion by 2018 up from $55 billion in 2013.

41 Telstra considers that applying price regulation to

this growing market risks inhibiting further development and innovation, which would not be in the LTIE.

4.2. Approach to pricing SMS termination 105. Telstra continues to consider that there is no requirement to set a regulated price for SMS

termination service. The nature of the SMS market in Australia, including the use of alternative messaging services and platforms, makes the application of regulatory assumptions used to set primary price terms for SMS termination highly uncertain. The limited number of jurisdictions (comparable or otherwise) from which to derive international benchmarks for SMS termination, as noted by the ACCC, suggests that regulated pricing of SMS termination is unnecessary. Further, while the approach taken by the ACCC in the Draft Decision is sound in principle, Telstra considers that the practical application of the approach by WIK-Consult is flawed.

106. As noted above, there are a very limited number of jurisdictions which regulate SMS

termination. For example, New Zealand set regulated SMS termination rates in 2011 based on a benchmark of three countries with TSLRIC frameworks – Denmark, Malaysia and Israel – while Indonesia bases regulated SMS termination rates on a LRIC cost model. SMS termination is not one of the defined markets in the European Union (EU) and at the end of 2014 only three of the 28 EU countries had elected to regulate SMS termination – Denmark, France and Poland. In December 2014 the French regulator, ARCEP, issued a draft determination in which SMS termination rates had been updated, however the European Commission found that there was insufficient evidence to support ARCEP’s findings that the SMS termination market had been analysed in accordance with European competition law principles for the purpose of ex ante

41

Juniper Research in http://www.businessreviewaustralia.com/marketing/1386/The-Growth-And-Benefits-Of-

Mobile-Messaging-Services-For-Business:-Part-One



PAGE 28

regulation.42

Subsequently ARCEP withdrew its draft decision but will continue to monitor the SMS

termination market.43

107. Given the limited number of jurisdictions from which to derive international benchmarks, WIK-

Consult devised an alternative approach for SMS termination, with the rate comprising two components:

a. The conveyance cost for SMS termination set relative to the mobile voice termination

rate based on the relative capacity used by the two services, i.e. a conversion factor. b. SMS-specific costs determined based on investment costs for SMS-specific network

elements, i.e. SMS centres. Conveyance cost

108. Telstra understands that WIK-Consult has derived the conveyance cost for SMS using normal design practice and industry accepted assumptions. However, the source of the assumption made regarding the proportions of 2G and 3G SMS in Australia is unclear. Network Strategies attempted to replicate the WIK-Consult calculations and concluded that WIK-Consult appear to have calculated the number of messages equivalent to a minute of voice assuming that the percentages of 2G and 3G SMS are the same as voice traffic – that is, 6% for 2G and 94% for 3G. Although WIK-Consult have stated that the proportions of voice traffic are based on actual data collected from the three MNOs in Australia, there is no information to support or justify adopting the same proportions for SMS. Further, this assumption affects the final calculations as the final conveyance cost per SMS is estimated using the voice termination rate (based on 2G and 3G blending of voice traffic). Telstra considers that, instead of directly using the termination rate for voice, the SMS cost calculation should use a blended rate based on the mix of 2G and 3G SMS.

SMS-specific costs

109. WIK-Consult obtained benchmark SMS Centre (SMSC) capital costs from the TSLRIC cost models of the benchmark sample used to determine mobile voice termination rates. This is despite requesting and receiving information from the three MNOs on the capital investments made into SMSCs. Telstra notes that the models used by WIK-Consult are up to three years old, and thus it is possible that the capital costs used are out of date. Further WIK-Consult does not appear to have given any consideration to the capacity of the SMSCs, which can influence costs.

110. Network Strategies compared the unit costs and other characteristics of SMSCs within the

cost models used by WIK-Consult. This is presented in

42

European Commission (2014), Commission decision concerning: Case FR/2014/1670: Wholesale SMS

termination on individual mobile networks Opening of Phase II investigation pursuant to Art.7 of Directive

2002/21/EC1 as amended by Directive 2009/140/EC, C(2014) 9270 final, 28 November 2014. 43

ARCEP (2015), ARCEP places SMS termination markets under supervision, press release, 29 January 2015.



PAGE 29

111. Table 3 below.



PAGE 30

Table 3: SMSC Characteristics

Network

Strategies unit

capital cost

(LCU 2015,

nominal)

Network

Strategies unit

capital cost

(AUD 2015)

WIK unit

capital cost

(AUD)

Capacity

(SMS per

second)

Maximum

utilisation

(%)

Denmark (2012) 1,227,166 285,930 330,920 400 38%

Denmark (2015) 1,194,834 278,396 - 400 38%

Mexico 2,463,595 5,092,251 2,930,945 1000 (HW)

400 (SW)

80%

Netherlands 1,705,336 3,161,693 2,788,014 1000 (HW)

400 (SW)

80%

Norway 15,256,452 3,112,316 4,255,727 500 70%

Portugal (2012)1

1,318,384 2,987,458 2,381,992 4500 (HW)

1500 (SW)

80%

Portugal (2015)1

1,317,348 2,985,111 - 4500 (HW)

1500 (SW)

80%

Romania 616,698 1,956,783 1,041,527 361 80%

Spain 432,526 907,440 854,998 1,000,000 -

Sweden2

3,163,714 578,960 1,395,360 1,000 40%

United Kingdom (2014) 23,406 54,442 5,229,858 5,800 80%

United Kingdom (2015) 24,895 57,906 - 5,800 80%

112. Although Network Strategies was not able to reconcile the costs in the models with those reported by WIK-Consult,



PAGE 31

113. Table 3 shows the wide variation in unit capacity and assumed maximum utilisation. WIK-

Consult also made a number of assumptions regarding economic life, operating costs and a common costs mark-up to derive an annualised SMSC cost. Given the information variability and assumptions made, Telstra considers that it would be preferable to use financial data sourced from Australian MNOs rather than a combination of benchmark costs and (seemingly unsupported) assumptions to avoid unnecessary subjectivity and the risk of regulatory error.

114. Telstra notes that WIK-Consult justified the use of information on investment in SMSCs from the cost models used for benchmarking mobile voice termination on the basis of “...a large disparity in the information provided by MNOs in terms of both the investment cost and the number of SMSCs.”

44 Although Telstra has no visibility of this disparity, Telstra considers that there was

sufficient time for WIK-Consult to work with the three MNOs to determine the source of any discrepancy and develop a consistent approach to the provision of SMSC cost information. It seems inappropriate to use information from alternate jurisdictions when Australian MNOs are able to provide information that reflects the efficient cost of providing SMS termination services in Australia. Telstra considers that the absence of specific guidance regarding the provision of information relating to investment in SMSCs is the likely cause of such disparity and should be revisited by the ACCC before a Final Decision is made regarding a regulated rate for SMS termination.

4.3. Draft SMS termination rates for 2016-2019

115. The ACCC has made a draft decision to adopt a regulated SMS termination rate of 0.03 cents

per SMS, which would be a flat rate to apply for the entire FAD period. As with the implementation of the proposed mobile voice termination rate, the ACCC considers it appropriate to provide a short period of transition for industry to adjust their commercial arrangements to reflect this change. The ACCC’s draft decision is that the SMS termination rate will come into effect from 1 January 2016.

116. As stated above, Telstra does not consider that it is necessary to set a regulated rate for SMS

termination. The nature of the SMS market in Australia suggests that regulated pricing is unnecessary and inappropriate given the wide range of substitutes for SMS and competitiveness of the provision of SMS, which is unlimited in a significant majority of mobile plans. Further, the approach taken by WIK-Consult to estimate a regulated SMS termination rate is based on information from other jurisdictions and the application of assumptions which make it unlikely that the rate reflects the efficient cost of providing SMS termination in Australia.

117. Telstra’s position is that the ACCC should not adopt a regulated price for SMS termination but continue to monitor the market as appropriate. Where there is evidence of market failure or cause for concern – which there currently is not – then the ACCC has the ability to set a price for the declared service.

44

ACCC, Draft Decision, Section 4.2.2



PAGE 32

05 FIXED-TO-MOBILE PASS THROUGH

118. The ACCC has made a draft decision not to include a mandated FTM pass-through

mechanism in the FAD. This position is premised on the following:

a. Evidence of FTM pass-through of past MTAS reductions and that further reductions in the mobile voice termination rate are expected to be passed on to end-users in the form of lower retail prices.

b. Additional intervention, in the form of a mandated FTM pass-through, is not likely to be in the LTIE.

Telstra agrees with the ACCC’s draft decision not to include a mandated FTM pass-through mechanism. As set out in previous submissions, Telstra considers that a mandated FTM pass-through mechanism would distort competition which would not be in the LTIE. Further, Telstra has provided evidence to the ACCC that reductions in MTAS rates have been more than passed through to customers and there is no evidence to suggest that future reductions in MTAS rates will not be similarly passed through.

5.1. A mandated FTM pass-through mechanism will distort competition

119. The draft decision sets out the ACCC’s view that imposing restrictions on an integrated operator in its retail pricing may have unintended consequences that do not promote the LTIE. Specifically, that a pass-through mechanism that only focuses on the reduction in prices may restrict a service provider’s ability to flexibly determine how it chooses to pass on its cost savings and limit (or even negate) potential improvements in the quality and range of retail services.

120. Telstra agrees with the ACCC’s position. A pass-through mechanism is likely to distort competition because it will force parties to pass through MTAS price reductions to the FTM price only and thus prevent any reductions from being passed through to other components of the fixed services bundle. Telstra has previously submitted evidence to the ACCC showing that the benefit to customers of a reduction in MTAS prices will be greater the more that those reductions are passed through on calling services other than FTM calls.

45 Telstra therefore considers that

imposing a pass-through requirement on integrated operators would harm the ability to meet customer needs, which would not be in the LTIE.

5.2. Reductions in MTAS rates have been passed through to end-users

121. The ACCC has undertaken analysis to estimate the extent of FTM pass-through that has occurred to date. The Draft Decision correctly identifies that it is difficult to ascertain precisely how the reductions in the mobile voice termination rate are reflected in the changes in retail prices of fixed line voice services. Telstra considers that this difficulty has contributed to the erroneous and misleading claims made by other stakeholders that Telstra has benefited (through a ‘windfall gain’) from reductions in the MTAS to the detriment of consumers. Nevertheless, the analysis undertaken by the ACCC and Telstra demonstrates that reductions in MTAS rates have been passed through to end-users.

122. In September 2014, Telstra submitted updated analysis which showed the steady decline of the FTM rate in supplying voice bundled services. Further, between December 2004 and

45

See Telstra, Supplementary submission in response to the Commission Discussion Paper on Domestic Mobile

Terminating Access Service (MTAS), 19 September 2011



PAGE 33

December 2013, Telstra’s FTM yield declined by 11.3 cents per minute while the effective MTAS rate paid by Telstra reduced by only 7.6 cents per minute.

46 This analysis was supported by the

ACCC which compared the estimated savings Telstra had made from the MTAS reductions and the changes in FTM call revenue, adjusted for changes in volumes, over the same period. The ACCC found that Telstra appeared to have passed on more than the savings it made from the MTAS reductions.

47

123. In the Draft Decision, the ACCC widened its previous analysis to consider information from

Imputation Testing reports. This analysis compared the average per minute retail FTM call rates with the total unit cost of FTM calls and found that approximately 94 per cent of any cost reduction (from either MTAS or another source) was passed through to retail prices for FTM calls. There is no evidence to suggest that future reductions in MTAS rates will not be similarly passed through.

124. The ACCC analysis also addressed claims from other stakeholders that Telstra’s retail margins indicated a lack of FTM pass-through. Specifically, the ACCC considered that “...near full pass-through of MTAS reductions and other cost savings to retail FTM prices is statistically consistent with an increasing retail margin.”

48 Telstra considers that the above suggests that a

mandatory pass-through mechanism is unnecessary and would not be in the LTIE.

46

See Telstra Corporation, Response to the Commission’s Mobile Terminating Access – Final Access

Determination Discussion Paper, September 2014. 47

ACCC, Domestic Mobile Terminating Access Service Declaration Inquiry: ACCC’s Final Decision, June 2014. 48

ACCC Draft Decision.



PAGE 34

06 DURATION OF REGULATED TERMS AND CONDITIONS 125. The ACCC has reached a draft decision that the regulated price terms and non-price terms in

the MTAS FAD should expire at the same time as the current MTAS declaration on 30 June 2019. However, the ACCC may review the price terms of the FAD before the expiry date should there be significant changes in circumstances which warrant an inquiry to vary the terms of the FAD. The ACCC refers specifically to the commercial launch of VoLTE on 4G networks as a potential reason to review the price terms of the FAD if there is evidence that regulated termination rates no longer reflect the efficient costs of providing the service in Australia.

126. Telstra agrees with the ACCC draft decision that the price and non-price terms of the MTAS FAD should expire at the same time as the current MTAS declaration on 30 June 2019. As noted elsewhere in this submission, Telstra considers that the timing of VoLTE deployment is uncertain at this point in time and unlikely to impact on the cost of providing the service during the duration of the FAD. Further Telstra considers that the ACCC should place priority on regulatory certainty at a time when the industry as a whole is going through a period of transition.

TELSTRA CORPORATION LIMITED submission_9.pdf · Telstra Corporation’s Response to the Commission’s Domestic Mobile Terminating Access Service Declaration Inquiry – Report of

Documents